We have shown that agent-based modeling and complex systems concepts can be practical and effective tools for middle school science and high school physics learning (Klopfer, 2009). This study builds on that experience, extending the use of agent-based models and complex systems simulations from the physical sciences to the secondary school biology curriculum. To this end, we designed and implemented a sequence of interactive off-computer and agent-based model building activities that would enable students to experience and experiment with the mechanisms that drive the emergence of large-scale global phenomena from smaller scales of agent actions and interactions. By including the act of model building (in many cases through computer programming) in the learning experience, students were able to participate in the full spectrum of interactions with simulations (Klopfer, 2009). Our objective here was to determine whether these modeling activities would provide the right prior knowledge (Schwartz, 2007), that when coupled with appropriate scaffolding and learning resources, would help students overcome misconceptions and build a robust understanding of evolutionary processes

Analysis of early student programming strategies revealed a need for instruction in systematic design and problem solving skills. The trial and error process employed by most students was inefficient, but it was good enough to help them discover an important complex systems principle; that in some systems, small code (behavior) changes could yield dramatic changes in the resulting systems-level patterns. To assess how the concrete act of programming affected the understanding of emergence, students were asked to explain how patterns in their simulations happened. While it has been reported that a coherent understanding of complex systems eludes most students (Jacobson & Wilensky, 2006), more than 85% of the students in this study were eventually able to make connections across scale by describing how population patterns emerged from individual agent behaviors and interactions. Post-activity written and oral discussions revealed that the simulation experience proved to be the right prior knowledge to account for evolution.